Most product strategy documents are not strategies
A strategy is a diagnosis, a guiding choice, and a coherent set of actions that follow from it. Most documents labelled product strategy contain none of these. They contain an ambition — 'become the leading platform for X' — a set of themes, and a roadmap. Ambition is not diagnosis. Themes are not choices.
The test is simple: does the document say what the company will not do? If the strategy is compatible with every plausible course of action, it cannot guide any decision, and it will be ignored by everyone below the leadership team within a month.
The second test: is there a falsifying condition? A real strategic bet can be wrong. If you cannot describe the evidence that would make you change course, you have not made a bet — you have described a hope.
The diagnosis is the hard part
Almost every strategy engagement begins with the wrong problem statement. The founder says growth has stalled and the roadmap needs sharpening. Three days of evidence gathering reveals that activation is fine, retention is fine, and the actual issue is that the highest-value segment churns at renewal because a competitor solved an adjacent problem they consider part of the same job.
Getting to that requires talking to people outside the building. Customer interviews, yes, but also lost-deal conversations, churned-customer conversations and the sales calls where the prospect politely disengaged. Internal opinion is systematically biased toward the problems the team already knows how to solve.
Once the diagnosis is right, the strategic choice is usually obvious and frequently uncomfortable, because it involves stopping work that people are emotionally invested in. That is why the diagnosis has to be evidenced rather than asserted — the evidence is what makes the stopping possible.
Turning strategy into a roadmap that survives
- —Every roadmap item names the outcome it is meant to move, not just the thing being built.
- —Three horizons: committed for the current quarter, likely for the next two, directional beyond. Precision beyond six months is theatre.
- —Each item carries the evidence that would prove it worked, agreed before the build starts.
- —An explicit 'not doing' list, published, with the reasoning. This is what makes the roadmap defensible when a big customer asks.
- —One page. If the strategy does not fit on a page, the team will not carry it in their heads, and a strategy nobody remembers is not operating.
The operating model is where strategy usually dies
A good strategy with a bad operating model produces a company that agrees on direction and drifts anyway. The operating model is the boring machinery: who decides what without escalating, what the planning cadence is, which metrics get reviewed weekly versus quarterly, and how discovery work gets funded when delivery pressure is high.
The most common failure is that discovery has no protected time. Everyone agrees that talking to customers matters, nobody has it in their calendar, and within two quarters the roadmap is being built from internal guesses again. The fix is structural, not motivational: a fixed number of customer conversations per person per month, reviewed like any other commitment.
The second most common failure is unclear decision rights. When nobody knows whether a PM can change scope without the founder, everything escalates, the founder becomes the bottleneck, and the strategy stops guiding anything because every decision is made ad hoc in the moment.